• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.23 no.11 s.170
• /
• pp.1929-1939
• /
• 1999

This paper presents a neural network approach, which was named PRONET, to 3D object recognition and pose calculation. 3D objects are represented using a set of centroidal profile patterns that describe the boundary of the 2D views taken from evenly distributed view points. PRONET consists of the training stage and the execution stage. In the training stage, a three-layer feed-forward neural network is trained with the centroidal profile patterns using an error back-propagation method. In the execution stage, by matching a centroidal profile pattern of the given image with the best fitting centroidal profile pattern using the neural network, the identity and approximate orientation of the real object, such as a workpiece in arbitrary pose, are obtained. In the matching procedure, line-to-line correspondence between image features and 3D CAD features are also obtained. An iterative model posing method then calculates the more exact pose of the object based on initial orientation and correspondence.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2019.10a
• /
• pp.1101-1104
• /
• 2019

키넥트는 RGBD 카메라로 인체의 뼈대와 관절을 3D 공간에서 스켈레톤 데이터수집을 가능하게 해주었다. 스켈레톤 데이터를 활용한 행동 분류는 RNN, CNN 등 다양한 인공 신경망으로 접근하고 있다. 본 연구는 키넥트를 이용해서 Skeleton Joints를 수집하고, DNN 기반 스켈레톤 모델링 학습으로 행동을 분류한다. Skeleton Joints Processing 과정은 키넥트의 Depth Map 기반의 Skeleton Tracker로 25가지 Skeleton Joints 좌표를 얻고, 학습을 위한 전처리 과정으로 각 좌표를 상대좌표로 변경하고 데이터 수를 제한하며, Joint가 트래킹 되지 않은 부분에 대한 예외 처리를 수행한다. 스켈레톤 모델링 학습 과정에선 3계층의 DNN 신경망을 구축하고, softmax_cross_entropy 함수로 Skeleton Joints를 집는 모션, 내려놓는 모션, 팔짱 낀 모션, 얼굴을 가까이 가져가는 모션 해서 4가지 행동으로 분류한다.

• Journal of the Korean GEO-environmental Society
• /
• v.14 no.9
• /
• pp.31-37
• /
• 2013

In this study installation diameter, interval, area replacement ratio and ground hardness of applicable ground in C.G.S method should be mastered through surrounding ground by conducting modeling. Optimum artificial neural network was selected through the study of the parameter of artificial neural network and prediction model was developed by the relationship with numerical analysis and artificial neural network. As this result, C.G.S pile settlement and ground settlement were found to be equal in terms of diameter, interval, area replacement ratio and ground hardness, presented in a single curve, which means that the behavior pattern of applied ground in C.G.S method was presented as some form, and based on such a result, learning the artificial neural network for 3D behavior was found to be possible. As the study results of artificial neural network internal factor, when using the number of neural in hidden layer 10, momentum constant 0.2 and learning rate 0.2, relationship between input and output was expressed properly. As a result of evaluating the ground behavior of C.G.S method which was applied to using such optimum structure of artificial neural network model, is that determination coefficient in case of C.G.S pile settlement was 0.8737, in case of ground settlement was 0.7339 and in case of ground heaving was 0.7212, sufficient reliability was known.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.15 no.1
• /
• pp.69-74
• /
• 2005

In this paper, we suggested user assistant soft computing method for 3D effect optimization. In order to maximize 3D effect of image, intervals among cameras have to be set up properly according to distance between cameras and an object. Two data such as interval and distance was obtained to use in neural network as the data for learning. However, if the data for learning was obtained by only human's subjective views, it could be that the obtained data was not optimal for learning because the data had an accidental ewer To obtain optimal data lot learning, we added candidature data to obtained data through data analysis, and then selected the most proper data between the candidature data and the obtained data for learning in neural network. Usually, 3D effect of image was affected by both distance from an object to cameras and an object size. Therefore, we suggested fuzzy inference model which was able to represent two factors like distance and size. Candidature data was added by fuzzy model. In the simulation result, we verified that the mote the obtained data was affected by human's subjective views, the more effective the suggested system was.

• KIPS Transactions on Software and Data Engineering
• /
• v.8 no.9
• /
• pp.379-390
• /
• 2019

In this paper, we propose a novel deep neural network model for Vision-and-Language Navigation (VLN) named LVLN (Landmark-based VLN). In addition to both visual features extracted from input images and linguistic features extracted from the natural language instructions, this model makes use of information about places and landmark objects detected from images. The model also applies a context-based attention mechanism in order to associate each entity mentioned in the instruction, the corresponding region of interest (ROI) in the image, and the corresponding place and landmark object detected from the image with each other. Moreover, in order to improve the success rate of arriving the target goal, the model adopts a progress monitor module for checking substantial approach to the target goal. Conducting experiments with the Matterport3D simulator and the Room-to-Room (R2R) benchmark dataset, we demonstrate high performance of the proposed model.

• Journal of IKEEE
• /
• v.25 no.3
• /
• pp.540-543
• /
• 2021

In this paper, we present an experimental comparative study of recurrent neural network based techniques for trajectory prediction of flight vehicles. We defined and investigated various relationships between input and output under the same experimental setup. In particular, we proposed a relationship based on the relative positions of flight vehicles. Furthermore, we conducted an ablation study on the network architectures and hyperparameters. We believe that this comprehensive comparative study serves as a reference point and guide for developers in choosing an appropriate recurrent neural network based techniques for building (flight) vehicle trajectory prediction systems.

• Journal of the Korea Society of Computer and Information
• /
• v.27 no.4
• /
• pp.27-36
• /
• 2022

In this paper, to improve the conditions of the fish processing site, we propose a method to predict the cutting point of fish according to the target weight using AI machine vision. The proposed method performs image-based preprocessing by first photographing the top and front views of the input fish. Then, RANSAC(RANdom SAmple Consensus) is used to extract the fish contour line, and then 3D external information of the fish is obtained using 3D modeling. Next, machine learning is performed on the extracted three-dimensional feature information and measured weight information to generate a neural network model. Subsequently, the fish is cut at the cutting point predicted by the proposed technique, and then the weight of the cut piece is measured. We compared the measured weight with the target weight and evaluated the performance using evaluation methods such as MAE(Mean Absolute Error) and MRE(Mean Relative Error). The obtained results indicate that an average error rate of less than 3% was achieved in comparison to the target weight. The proposed technique is expected to contribute greatly to the development of the fishery industry in the future by being linked to the automation system.

• Tunnel and Underground Space
• /
• v.19 no.3
• /
• pp.203-212
• /
• 2009

The prediction of near field behavior around an underground high-level radioactive waste repository is important for the repository design as well as the safety assessment. In this study, a sensitivity analysis for seven parameters consisted of design parameters and material properties was carried out using a three-dimensional finite difference code. From the sensitivity analysis, it was found that the effects of borehole spacing, tunnel spacing, cooling time and rock thermal conductivity were more significant than the other parameters. For getting a statistical distribution of buffer and rock temperatures around the repository, an artificial neural network, backpropagation, was applied. The reliability of the trained neural network was tested with the cases with randomly chosen input parameters. When the parameter variation is within ${\pm}10%$, the prediction from the network was found to be reliable with about a 1% error. It was possible to calculate the temperature distribution for many cases quickly with the trained neural network. The buffer and rock temperatures showed a normal distribution with means of $98^{\circ}C$ and $83.9^{\circ}C$ standard deviations of $3.82^{\circ}C$ and $3.67^{\circ}C$, respectively. Using the neural network, it was also possible to estimate the required change in design parameters for reducing the buffer and rock temperatures for $1^{\circ}C$.

• The Transactions of the Korean Institute of Electrical Engineers P
• /
• v.50 no.3
• /
• pp.138-144
• /
• 2001

• Journal of the Korea Society of Computer and Information
• /
• v.14 no.9
• /
• pp.11-19
• /
• 2009

A RAM-based Neural Net(RBNN) which has multi-discriminators is more effective than RBNN with a discriminator. Experience Sensitive Cumulative Neural Network and 3-D Neuro System(3DNS) that accumulate the features point improved the performance of BNN, which were enabled to train additional and repeated patterns and extract a generalized pattern. In recognition process of Neural Net with multi-discriminator, the selection of class was decided by the value of MRD which calculates the accumulated sum of each class. But they had a saturation problem of its memory cells caused by learning volume increment. Therefore, the decision of MRD has a low performance because recognition rate is decreased by saturation. In this paper, we propose the method which improve the MRD ability. The method consists of the optimum MRD and the matching ratio prototype to generalized image, the cumulative filter ratio, the gap of prototype response MRD. We experimented the performance using NIST database of NIST without preprocessor, and compared this model with 3DNS. The proposed MRD method has more performance of recognition rate and more stable system for distortion of input pattern than 3DNS.